SUBSTRATE ANALOGUES

Like the entire subject of antimetabolites, this is a complex subject and is interrelated with many of the sections above. With B6-dependent en­

zymes it is peculiarly complex because of interreaction of substrate and substrate analogues with coenzyme. Inhibition by "mass action" is compli­

cated in interpretation because of this effect. Nevertheless, true com­

petitive inhibition by substrate analogues occurs, and for this reason this section is a highly important one. The subject is discussed in Volume I I by Sourkes and DTorio.

A. Bacterial and Plant Decarboxylases

1. GLUTAMIC ACID

Glutamic acid decarboxylation by whole cells or extracts of several species of bacteria is inhibited by aspartic acid (Storck, 1951). There is little or no effect on decarboxylation by plant extracts of other amino acids (Okunuki, 1937; Schales and Schales, 1946b).

2. HISTIDINE

Histidine decarboxylation by E. coli is unaffected by benzoylhistidine or acetylhistidine (Geiger, 1944). Plant histidine decarboxylase is not inhibited by D-histidine, unlike the mammalian enzyme (vide infra) (Werle and Raub, 1948).

9. INHIBITION OF AMINO ACID DECARBOXYLASES 357

3. TYROSINE

Tyrosine decarboxylation by S. faecalis preparations is not inhibited by 23 tyrosine analogues tested (Epps, 1944; McGilvery and Cohen, 1948;

Frieden et al, 1951; Lestrovaya and Mardashew, 1960), even if present in amounts exceeding the substrate eightfold (McGilvery and Cohen, 1948).

Blaschko and Stiven (1950) showed that o-, m-, and p-chlorophenylalanines act neither as substrates nor inhibitors of tyrosine decarboxylase in S.

faecalis.

B. Mammalian (and Fowl) Decarboxylases in vitro (cf. also Section XVI, G)

1. CYSTEIC ACID

Cysteic acid decarboxylase is not inhibited by asparagine (Werle and Bruninghaus, 1951), cysteine, phosphoserine, aspartic acid, or benzene-sulfinic acid, but is inhibited 76% by 10~

3

M cysteinesulfinic acid (Simon-net et al., 1960) and by homocysteic acid, a nonsubstrate (Blaschko, 1945b).

2. GLUTAMIC ACID

Glutamic acid decarboxylase of brain homogenates is competitively in­

hibited by p-hydroxyphenylacetic > phenylpyruvic > p-hydroxyphenyl-pyruvic acids, but not by phenylacetic acid or phenylalanine (Hanson, 1958) (cf. Sections X V I , C and H below). Greig et al. (1959) reported that several derivatives of α-methyltryptamines inhibited (cf. Section X V I , I below, and substrate-coenzyme interaction, discussed by Sourkes and DTorio in Volume I I of this treatise).

3. HISTIDINE

Histidine decarboxylase (nonspecific) is not inhibited by other amino acids, with the exception of cysteine; equivocally by tyrosine and trypto­

phan; dopa, which strongly inhibits (Werle and Heitzer, 1938; Werle and Koch, 1949; Ganrot et al., 1961); and α-methyldopa (but cf. Section X V I , I below). Interestingly, D-dopa inhibits more than L-dopa, according to Werle (1941). Adrenaline and noradrenaline, D-histidine, imidazole, and benzylimidazole (Priscol) also inhibited. Beiler et al. (1949) confirmed the lack of effect of other amino acids, but found mild inhibition by 1.0 mg/ml iV-sulf anilyl-4-aminobenzimidazole, benzoxazolone, 3-benzothioph ene-a-aminopropionic acid, and β-2-thienylalanine. Schayer and Kobayashi

(1956) and Schayer (1956b) confirmed the latter with histidine decar­

boxylase of rabbit blood platelets and rat peritoneal mast cells, and ex­

tended the list to β-3-thienylalanine, methylhistidine, tyrosine, tryptophan and 5-HTP. Phenylalanine, D-histidine, histamine itself, and phenylalanine

were inactive, and thiolhistidine only slightly. Caffeic acid and catechol also inhibit histidine decarboxylase of rabbit kidney, but these agents (and α-methyldopa) do not inhibit the specific L-histidine decarboxylase of embryonic rat liver (Ganrot et al., 1961).

C. Mammalian Dopa Decarboxylase in vitro

This subject, as well as dopa decarboxylase inhibition in vivo, has been adequately reviewed by Clark (1959), Clark and Pogrund (1961), and Sourkes and DTorio (see Volume I I ) and will not be repeated here in extenso.

Drell (1957) incubated tyrosine-C 14

with beef adrenal slices and found that the decarboxylase inhibitor, 5-(3-hydroxylcinnamoyl)salicylate (Clark, 1959; Clark and Pogrund, 1961), markedly decreased the incorporation of label into the catechol amine fraction and increased that in the catechol acid fraction, which includes dopa.

Fellman (1959) found that 5-HTP competitively inhibits dopa decar­

boxylation. During purification, the ability to cleave dopa, o-tyrosine, and 5-HTP remained constant at each step, leading him to conclude that the same enzyme is involved. Rosengren (1960) confirmed this and, in addi­

tion, found that dopa and 5-HTP decarboxylations are cross-inhibited also by ra-tyrosine, o-tyrosine, and caffeic acid, the latter in confirmation of Hartman et al. (1955). Werle and Aures (1960) also found inhibition by caffeic acid, chlorogenic acid, and 5-(3-hydroxycinnamoyl)salicyclic acid, in confirmation of Hartman et al. Griesemer et al. (1961), in a preliminary abstract, recently extended the studies of Hartman et al. (1955) to add 45 more compounds to their original list of over 200 compounds. Sixteen of the 45 gave good inhibition, and the essential structural requirements formerly elucidated were confirmed and extended, the best inhibitors having the structure R—CH=CH—CO—R', where R = 3-hydroxyphenyl, 3,4-dihydroxyphenyl, or 5-hydroxyindole, and R ' = OH, O-alkyl or aryl.

D, Mammalian 5-HTP Decarboxylase in vitro

C. J. Clark and associates (1954) found no inhibition of this enzyme by tryptophan, 7-hydroxytryptophan, or 5-HT up. to 10~

2

M. 5-Benzoxy-tryptophan inhibited at 10~*

3

M. The list was extended in the same labora­

tory (Fréter et al., 1958) to include 10 other 5-HT and 5-HTP analogues, the most effective of which was 2,5-dihydroxytryptophan, which gave 68% inhibition at concentrations equimolar with substrate. Ozaki (1959) reported that iV-methyldopa inhibits 5-HTP decarboxylase of rat brain.

9. I N H I B I T I O N OF A M I N O ACID DECARBOXYLASES 359 Yuwiler et al. (1959, 1960) examined the best dopa decarboxylase in­

hibitors reported by Hartman et al. (1955) on 5-HTP decarboxylase and found them equally active and competitive. In addition, they found com­

petitive inhibition by the analogue l-[5-hydroxyindolyl-3]-2-(3-carboxy-4-hydroxylbenzoyl)ethylene. This compound is also a good inhibitor of dopa decarboxylase in vitro (cf. also Griesemer et al., 1961). /3-(3-Indolyl)-acrylic acid was a fair inhibitor, but 5-(3-indoleacryloyl)-salicylic acid and l-[5-oxyindolyl-(3)]-butenone-(3) were inactive, which is interesting since the m-hydroxyphenyl analogue of the latter, with the same side chain,

—CH=CH—CO—CH3 (m-hydroxybenzalacetone), inhibits dopa decar­

boxylase (Hartman et al., 1955).

Recently, Erspamer et al. (1961) examined 22 tryptophan analogues as substrates of 5-HTP decarboxylation by guinea pig kidney extracts and reported marked inhibition at pH 6.8 by 8 μηιοΐββ of caffeic acid and 1,4-bis-(3,4-dihydroxycinnamoyl)quinic acid (1,4-dicaffeylquinic acid, "eyna-rine"). Inhibition of decarboxylation of 5-HTP was more complete than that of 4-HTP (the only other analogue of the 22 examined which was a substrate) or of dopa.

Ε. Glutamic Acid Decarboxylase in vivo

This reviewer could find no work on this subject except some yielding indirect evidence (cf. Section X V I , I below).

F. Dopa Decarboxylase in vivo (cf. Clark, 1959; Clark and Pogrund, 1961;

and Sourkes and D'lorio, Volume II)

The first clear-cut evidence of dopa decarboxylase inhibition by dopa analogues in vivo which also are active in vitro was presented in abstract form by Pogrund and Clark (1953) and Pogrund et al. [1955; cf. Clark and Pogrund (1961), which describes methods in detail]. In a review on dopa decarboxylase inhibitors, Clark (1959) listed 30 active compounds and 50 inactive analogues. In the meantime, Dengler and Reichel (1957) and Westermann et al. (1958) demonstrated the inhibition of dopa, dops, and 5-HTP in vivo with a-methyldopa.

During this work, it was noted that if a Lineweaver-Burk analysis is made in vivo (for discussions of this, cf. Chen and Russell, 1950; Gaddum, 1957; Nickerson, 1959), the inhibitors appear to be competitive (Clark, 1959; Clark and Pogrund, 1961). It was also found that there is no strict parallelism between the inhibition in vitro and in vivo, but those substances which were inactive in vitro almost always were also inactive in vivo. All

substances tested were rapidly metabolized, and the inhibitory effects disappear within 30-60 minutes, including α-methyldopa (vide infra).

Because of this, further pharmacological, physiological, and clinical studies were not planned until compounds with more prolonged effects could be developed. In a few experiments (Clark and Pogrund, 1961), however, resynthesis of catechol amines by the insulin-depleted adrenal glands of rats seemed suppressed by repeated injections of two inhibitors which later were shown to be one-fifth as active as the best one, 5-(3-hydroxycinnamoyl)salicylic acid. Brodie et al. (1962), Drain et al. (1962), and Burkard et al. (1962), however, showed that nearly complete inhibition of 5-HTP/dopa decarboxylase in vivo by α-methyldopa, a-benzyloxy-amine, and α-benzylhydrazine do not affect the endogenous levels of catechol amines and 5-HT, possibly because only a small fraction of the nonrate-limiting decarboxylase activity available in normal tissues is sufficient for a physiological rate of decarboxylation.

Indirect evidence of cross competition of dopa and 5-HTP decarboxyla­

tion was recently afforded by Kato (1959), who showed that 5-HTP po­

tentiation of barbiturate hypnosis in mice is completely reversed by dopa.

G. Quinones and Potential Quinoids

In addition to reacting with thiol groups, such compounds also may react with free amino groups. Many compounds described in the literature as inhibitors of amino acid decarboxylases may belong to this category rather than to those claimed by the authors. If compounds have potential quinoid structures, precautions should be taken to prevent oxidation to quinones by adding, e.g., cysteine to substrate and glutathione to enzyme before reacting under anaerobic conditions (even then, such reductants have not prevented quinone formation in some cases). If this and other precautions, such as performing a Lineweaver-Burk analysis, have not been taken, re­

ports of inhibition by "substrate competition/

1

"displacement," or by

"metabolic antagonism" of various physiological functions in vivo are open to criticism. This is particularly true in the clinical literature. Examples are the clinical reports on the oral administration of flavonoids ("vitamin P " ) in allergies, based on the report that they were shown to inhibit histamine synthesis in vitro. To the reviewer's knowledge, the first ex­

amples which demonstrated unequivocally the alteration of a clinical entity by inhibiting an amino acid decarboxylase in human subjects are the recent reports of Oates, Gillespie, Sjoerdsma, Crout, and Udenfriend at the National Institutes of Health, Bethesda, Maryland (cf. Section X V I , I below).

9. I N H I B I T I O N OF A M I N O ACID DECARBOXYLASES 361 Bacterial decarboxylation of basic amino acids is inhibited by tannins and their precursors, such as o-dihydroxyphenolearboxylic acids and poly­

phenols, the former being more effective (Kimura et al, 1958). The in­

hibition is not antagonistic to the coenzyme, is noncompetitive with the substrate, and is not reversed by cysteine. Inhibition of mammalian dopa and histidine decarboxylases in vitro has been reported for many quinones and potential quinoid compounds, including quinone, hydroquinone, catechol, pyrogallol, dopa, catechol amines, and o-dihydroxyphenols in general, including flavonoids, anthocyanins, hematoxylin, and related compounds (Werle, 1941; Werle and Koch, 1949; Malkiel and Werle, 1951; Werle and Aures, 1960; Imiya, 1941; Martin et al, 1942, 1949;

Martin, 1951; Bargoni, 1946; Gonnard, 1951; Gâbor et al, 1952a, b; Parrot and Reuse, 1954; Hartman et al., 1955; Kimura et al., 1958), and glutamic acid decarboxylase by adrenochrome (Holtz and Westermann, 1956a, b, 1957). The latter may not be a purely nonspecific quinone inhibition, however, since Deltour et al. (1959a, b) found activation of the same enzyme by adrenochrome derivatives in which the quinone function is blocked.

Schayer et al. (1955) found no inhibition of mammalian histidine de­

carboxylase by cf-catechin. Hartman and co-workers (1955) examined a series of flavonoids for their ability to inhibit dopa decarboxylase, includ­

ing flavones, flavans, flavanones, chalcones, coumarins, and related com­

pounds, and discuss structural requirements. The list was extended recently in a preliminary abstract by Griesemer et al. (1961) to include additional analogues. If cysteine is added to the side arm of the Warburg vessel with substrate, and glutathione with enzyme before mixing to start the reaction, many potential quinoid structures had markedly less inhibition, and a Lineweaver-Burk analysis showed the inhibition was competitive. With­

out such precautions, the inhibition was noncompetitive or pseudocom-petitive. Incubation of inhibitor with enzyme prior to starting the reaction markedly enhanced inhibition. Compounds with 3,4-dihydroxy groupings in the flavone type (or the analogous 3',4'-dihydiOxy groupings in the chalcones) were most active, but the 3-hydroxychalcones, which are non-quinoid, also were quite active.

H. Ketonuria

Weil-Malherbe (1955) reported low blood adrenaline levels in mentally defective patients including phenylketonurics (phenylpyruvic oligophrenia).

This led Fellman (1956) to examine the effects of aromatic acids associated with the disease on dopa decarboxylation of beef adrenal medullary ex­

tracts. Phenylalanine had no effect, but phenylpyruvic, phenyllactic, and

phenylacetic acids, in that order, inhibited at concentrations from 3 to 30

^mole/ml, depending on the compound. The author speculated on whether this might be one reason for Weil-Malherbe's observation. These results were surprising in view of the results of Hartman et al. (1955), who found no appreciable effects of these compounds. However, Davison and Sandler (1958) confirmed the effect of these compounds on 5-HTP decarboxylase in vitro, except that phenylacetic acid was more active than either phenyl-pyruvic or phenyllactic acid at the same concentrations.

Pare and associates (1957, 1958a, b) speculated that, in addition to a defective hydroxylation of phenylalanine in ketonurics, there might be a similar defect in 5-HT synthesis and found low blood 5-HT and urine 5-HIAA levels in such patients. Low phenylalanine diets increased the blood 5-HT in six of seven cases. 5-HTP tolerance tests in four cases showed subnormal urinary 5-HT and 5-HIAA excretion. However, therapy with 5-HTP was of no benefit.

Sandler (1959a, b; Sandler and Close, 1959) reported that phenylacetic acid administered in doses of 5 gm by mouth to five ketonurics decreased urinary 5-HIAA in three of them, although there was no relief of symptoms.

Phenylalanine had no such effect.

Hanson (1958, 1959) examined the inhibitory effect of such compounds on glutamic acid decarboxylation by brain in vitro and found p-hydroxy-phenylacetic > phenylpyruvic > p-hydroxyphenylpyruvic acids in con­

centrations of 25-100 /xmoles/ml. Phenyllactic acid and phenylalanine had no effect.

Baldridge et al. (1959) found that, following the decrease of serum phenylalanine seen in ketonurics on a low phenylalanine diet, urinary 5-HIAA excretion increased. Oral administration of tryptophan or 5-HTP produced the same effect. Huang et al. (1961) found that rats on high phenylalanine and tyrosine diets excrete 5-10 times more phenylpyruvic acid; this was correlated with decreased 5-HIAA excretion and lowered plasma phenylalanine and 5-HT. Thus, such diets simulate the condition seen in clinical ketonurics.

Tashian (1960) found increased urinary indoleacetic acid in normal sub­

jects fed phenylpyruvic, phenylacetic, or o-hydroxyphenylacetic acids, the latter being more effective. The administration of phenylpyruvic or o-hy-droxyphenylacetic acids caused an increase in 5-HIAA excretion which was less than that of indoleacetic acid. Assuming that the same metabolites which can inhibit 5-HTP decarboxylase can also inhibit tryptophan de­

carboxylase, Tashian postulated that inhibition of tryptamine and 5-HT synthesis, therefore, could shunt an increased amount of tryptophan through transanimation to indolepyruvic acid and IAA, thus accounting for the observations. Subsequently (1961), he found that glutamic acid

9. I N H I B I T I O N OF A M I N O ACID D E C A R B O X Y L A S E S 363 decarboxylation by rat brain homogenates and E. coli powder was com­

petitively inhibited by phenylpyruvic, p-hydroxyphenylpyruvic, phenyl-acetic, p-hydroxyphenylphenyl-acetic, o-hydroxyphenylacetic acids, and deriva­

tives of valine and leucine. He postulated that if compounds such as these, which are formed in greater amounts in phenylketonuria and branched-chain ketonuria (maple syrup disease), reach the developing brain, they might limit the formation of gaba and of amines possibly essential to normal neurological function.

I. a-Alkyl Substrate Analogues (see also Section IV)

This subject has been treated by Sourkes and DTorio in Volume I I . The subject was also briefly reviewed up to November, 1960 (Clark and Pogrund, 1961).

Pfister et al. (1955) and Stein et al. (1955) described the syntheses of some α-methyl homologues of glutamic acid, methionine, diaminopimelic acid, and phenylalanine, including those of tyrosine and dopa. These were prepared as potential antimetabolites, including amino acid decarboxylase antagonists.

Roberts (1952a, 1953) examined some of these and other glutamic acid analogues for their ability to inhibit glutamic acid decarboxylation by acetone powders of E. coli and mouse brain. Noninhibitors of the bac­

terial enzyme included the diastereoisomer racemates of a-hydroxy-glutamic acid, methionine sulfoxide, various iV-substituted α-amides of glutamic acid, and pyrrolidonecarboxylic acid analogues. Inhibition was best with α-hydroxyiminoglutaric and DL-a-methylglutamie acid, the former probably because of its hydrolysis to H O N H2 and the latter by competitive inhibition. Waksman (1957) found that α-methylglutamic acid inhibits glutamic acid decarboxylation by acetone powder of Torulopsis utilis.

Roberts (1952a) reported that α-methylglutamic acid itself was not a substrate, at least by the method used. It also was the most potent of the series in inhibiting the utilization of glutamic acid for Lactobacillus ara­

binosus growth. It was a weak to moderate inhibitor of the brain enzyme.

Evidence for inhibitory action in vivo was suggested by the observation that α-methylglutamic acid enhanced the incidence of audiogenic seizures in susceptible mice, which was counteracted by glutamic acid. Ginsburg and Roberts (1951) reported that, of the various enhancing agents, me­

thionine sulfoxide and α-methylglutamic acid aggravated the seizures in proportion to their potency as metabolic antagonists of glutamic acid in bacterial growth. Ginsburg (1954) demonstrated a seizure enhancement effect of α-methylglutamic acid, but enhancement was also induced by

other organic acids. Other examples of inhibitory effects of α-methyl metabolite analogues exist in the literature, such as the inhibition of B6-dependent active uptake of amino acids inhibition of the enzymic formation by α-methylglutamic acid of the transferase in sheep brain, which produces glutamohydroxamic acid from hydroxylamine and glu-tamine, and of glutaminase in dog kidney (Lichtenstein et al., 1953a, b).

Other antimetabolite effects of α-methyl amino acids, as well as the mech­

anism of action, were reviewed briefly by Umbreit (1955) (cf. also Christen­

sen, Section I V ) .

Sourkes (see Sourkes and DTorio, Volume I I ) tested a series of 22 phenylalanine analogues, including the α-methyl derivatives prepared by Pfister et al., on dopa decarboxylase in vitro. Among them, a-methyl-DL-dopa and α-methyl-DL-ra-tyrosine were good inhibitors at 1-5 Χ 10~

4

M, respectively, when preincubated with enzyme before adding substrate. Subsequently, Hartman et al. (1955) found that, when added simultaneously, the inhibition by a-methyldopa is only fair, 35% at 10~

3

M. When preincubated 15 minutes with enzyme, in contrast, 5-(3,4-dihydroxycinnamoyl)salicylate inhibited 100% at 10~

6

M (cf.

also Griesemer et al., 1961).

Simmonet and co-workers (1960) reported that DL-a-methylcysteic acid, 1.5 X 10~

3

M, inhibits cysteic acid decarboxylase of chick embryo tissue in vitro.

Greig et al. (1959, 1961) examined the inhibitory effects of relatively high concentrations (10~

2

M) of a series of α-alkyltryptamines on 5-HTP decarboxylase in vitro and in vivo and found inhibition in vitro by 5-hy-droxy-a-methyltryptamine-creatinine sulfate and α-methyltryptamine but not by α-ethyltryptamine (Etryptamine, Monase). The 5-hydroxy ana­

logue was most active. Iproniazid also inhibited at 5 X 10~

3

M. Both compounds also were active in intact mice (cf. Fréter et al., 1958 for other inhibitory 5-hydroxytryptamine analogues). Glutamic acid decarboxyla­

tion by brain in vitro was unaffected. The predominant action of these 5-HT analogues was, however, inhibition of monoamine oxidase. Yuwiler et al. (1959) did not confirm Greig et al. on an inhibitory effect of a-methyl-tryptamine (as the methanesulfonate) on 5-HTP decarboxylase in vitro.

Van Meter et al. (1960) found that α-methyltryptamine inhibited 5-HTP decarboxylation by brain in vivo, as did methyltryptamine to a lesser degree. These compounds also increased 5-HT levels in brain, .probably by blocking monoamine oxidase. Effects on binding-release mechanisms were not examined. The effects were correlated with electroencephalo-graphic and behavioral changes.

Weissbach et al. (1960a, 1961) and Lovenberg et al. (1962) showed that α-methyldopa inhibits decarboxylation by a semipurified guinea pig

9. I N H I B I T I O N OF A M I N O ACID D E C A R B O X Y L A S E S 365 kidney preparation, of 5-HTP, dopa, tryptophan, phenylalanine, tyro­

sine, and histidine and that several of the α-methyl analogues are them­

selves decarboxylated, in proportion to their inhibitory power. This shows that there need not be a hydrogen atom on the α-carbon for de­

carboxylation. They propose that the same enzyme is involved in all of these reactions and propose to term it the "general aromatic amino acid decarboxylase." Previously, they had claimed that 5-HTP and dopa decarboxylases were distinct enzymes (C. J. Clark et al, 1954) but re­

tracted this in agreement with the conclusions made by Yuwiler et al.

(1959, 1960), Werle and Aures (1959, 1960), Holtz (1959), Fellman et al.

(1960), and Rosengren (1960). The original discrepancy was attributed to differences in B6-P04 requirement with different substrates (Werle and Aures), and subsequent work by Lovenberg et al. (1962) showed the ratios of activity with different substrates remained constant when more highly purified enzyme preparations were used.

Erspamer et al. (1961) have confirmed the inhibitory effect of a-methyl-dopa on 5-HTP decarboxylase in vitro. Inhibition of the decarboxylation

Erspamer et al. (1961) have confirmed the inhibitory effect of a-methyl-dopa on 5-HTP decarboxylase in vitro. Inhibition of the decarboxylation

In document of 9 (Pldal 42-67)